A recent Wall Street Journal article, “Big Tech Accounting Creates a Blind Spot in the AI Boom,” highlights a problem that anyone analyzing hyperscaler financials knows intimately: depreciation expense is difficult to find.
Not because the companies don’t report it but the numbers are buried in their cash flow statements or in 10-K footnotes. The problem is where it shows up on the income statement. At Microsoft, Alphabet, Amazon, Meta, and Oracle, depreciation isn’t disclosed as a standalone line item. It’s allocated across COGS, R&D, and SG&A expenses. For investors trying to understand the true cost of the AI infrastructure buildout, this creates a genuine blind spot.
On November 4, 2024, the FASB issued ASU 2024-03, which requires disaggregated disclosure of income statement expenses for public business entities (PBEs). The ASU does not change the expense captions an entity presents on the face of the income statement; rather, it requires disaggregation of certain expense captions into specified categories in disclosures within the footnotes to the financial statements.
With these five companies projecting around $695 billion in combined capital expenditure for 2026, an incredible 83% increase from $379 billion in 2025, depreciation is no longer a minor figure. It’s arguably the most important non-cash expense on Big Tech’s income statement. And it’s set to grow significantly.
I decided to quantify this problem using Daloopa’s Scout MCP integration in Claude, pulling six years of data across all five hyperscalers in eminutes. Here’s what I discovered.
The Problem: Depreciation Spread Across Three Line Items
When Amazon spends $131.8 billion in capital expenditures (as it did in FY2025), those assets hit the balance sheet as property, plant & equipment. Over the following years, that cost flows through the income statement as depreciation, but not in one place.
Consider how each company handles it:
- Amazon reports depreciation of property and equipment within cost of sales (for fulfillment and AWS infrastructure) and within technology and content (for R&D-related assets). There is no single line item on the income statement that tells you total depreciation expense.
- Microsoft allocates depreciation into COGS, R&D and SG&A. The consolidated figure only appears in the cash flow statement’s D&A addback or in annual report footnotes.
- Alphabet combines depreciation with impairment charges in its disclosures, and distributes the cost across its operating expense categories.
- Meta embeds depreciation within cost of revenue and R&D, with the total only reconcilable through the cash flow statement.
- Oracle follows the same pattern, splitting depreciation across cost of services, R&D, and G&A.
For a fundamental analyst, this means you can’t simply glance at the income statement and know how much of operating income is being consumed by depreciation. You have to dig for it across filings, footnotes, and supplemental disclosures.
The Solution: Daloopa MCP in Claude
This is exactly the type of problem Daloopa was built to solve.
Using Daloopa’s MCP (Model Context Protocol) integration directly inside Claude, I was able to:
- Discover all five companies in Daloopa’s database with a single API call using ticker symbols
- Search for the exact depreciation series. Daloopa has already done the work of isolating total depreciation expense from the cash flow statement and footnotes, standardized across companies
- Pull six years of annual data (FY2020–FY2025) for depreciation, revenue, operating income, capital expenditures, and operating cash flow across all five companies
- Build a comparative analysis with computed ratios (depreciation as % of revenue, depreciation as % of operating income, capex as % of CFO) instantly
The entire process from initial query to a complete, interactive dashboard with seven chart views and five data tables took a single conversation in five minutes.
What the Data Reveals
Depreciation Is Exploding
Total depreciation expense across the five hyperscalers increased from $47.7 billion in FY2020 to $106.9 billion in FY2025, representing a 17.5% compound annual growth rate. And this is before the depreciation wave from 2024 and 2025’s record capital spending has fully hit.
| Company | FY2020 D&A | FY2025 D&A | 5Y CAGR |
|---|---|---|---|
| MSFT | $10.7B | $22.0B | 15.5% |
| GOOG | $12.9B | $21.1B | 10.3% |
| AMZN | $16.2B | $41.9B | 20.9% |
| META | $6.4B | $18.0B | 23.0% |
| ORCL | $1.4B | $3.9B | 22.9% |
Amazon alone recognized $41.9 billion in depreciation in FY2025. To put that in perspective, that’s larger than the total annual revenue of companies like Starbucks or Nike.
Depreciation as % of Operating Income: The Real Margin Story
This is where the analysis gets interesting. Depreciation as a percentage of operating income reveals how much of a company’s profitability is being consumed by infrastructure costs:
| Company | FY2020 | FY2025 | Change |
|---|---|---|---|
| AMZN | 70.7% | 52.4% | -18.3pp |
| ORCL | 9.9% | 21.9% | +12.0pp |
| META | 19.5% | 21.6% | +2.1pp |
| MSFT | 20.2% | 17.1% | -3.1pp |
| GOOG | 31.3% | 16.4% | -14.9pp |
A few standout observations:
Amazon’s depreciation accounted for 52.4% of its operating income in FY2025. In FY2022, during the post-pandemic margin compression, depreciation accounted for 203.7% of operating income, meaning Amazon’s infrastructure costs alone exceeded its operating profit. The company has improved, but depreciation is still its largest expense compared to profits.
Oracle’s ratio nearly tripled from 9.9% to 21.9% as the company pivoted aggressively toward cloud infrastructure (i.e. OCI). With FY2026 capex guidance of $50 billion (up from $21.2 billion in FY2025), this ratio is poised to rise significantly.
Microsoft and Alphabet demonstrated the strongest operating leverage. Despite extensive AI infrastructure expansions, both companies increased operating income quickly enough to reduce depreciation’s share. Microsoft’s 17.1% and Alphabet’s 16.4% are the lowest in the group, highlighting their ability to apply the AI investments to their high-margin core businesses, not just their cloud businesses.
The Capital Expenditure Tsunami
The depreciation story is really a forward-looking capex story. What companies spend on infrastructure today becomes depreciation expense over the next 5–6 years. And the 2026 guidance numbers are staggering:
| Company | FY2025 Capex | 2026E Guidance | YoY Growth |
|---|---|---|---|
| AMZN | $131.8B | ~$200.0B | +52% |
| GOOG | $91.4B | $175–185B | +97% |
| MSFT | $64.6B | ~$140.0B | +117% |
| META | $69.7B | $115–135B | +79% |
| ORCL | $21.2B | ~$50.0B | +136% |
| Total | $378.7B | ~$695B | +83% |
This $695 billion in 2026 capex will generate depreciation charges for years to come. Assuming a blended 5-year useful life, we’re looking at an incremental ~$139 billion in annual depreciation expense layering onto these companies’ income statements on top of the existing base.
Meta’s CFO explicitly warned investors that “higher depreciation” would be one of the primary drivers of expense growth in 2026. Alphabet similarly flagged that expenses would “meaningfully increase” due to depreciation and energy costs. These aren’t abstract accounting concepts anymore. They’re the dominant factor shaping margin trajectories.
Capex as % of Operating Cash Flow: The Free Cash Flow Squeeze
Perhaps the most alarming metric in the analysis: how much operating cash flow is being consumed by capital expenditure.
| Company | FY2020 | FY2025 | Change |
|---|---|---|---|
| ORCL | 11.9% | 101.9% | +90.0pp |
| AMZN | 60.7% | 94.5% | +33.8pp |
| META | 39.0% | 60.2% | +21.2pp |
| GOOG | 34.2% | 55.5% | +21.3pp |
| MSFT | 25.4% | 47.4% | +22.0pp |
Oracle is already spending more than 100% of its operating cash flow on capex, with free cash flow negative $10 billion in Q2 FY2026 alone. Barclays has warned the company could face liquidity pressure by late 2026. Its bonds, while still technically investment-grade rated, are trading at credit default swap spreads reminiscent of the 2009 financial crisis.
At 94.5%, Amazon generated virtually no free cash flow in FY2025 despite $139.5 billion in operating cash flow. With $200 billion in guided capex for 2026, Amazon will need to either grow operating cash flow proportionally or accept negative free cash flow.
The 2026E estimates suggest these ratios will worsen before they improve with Oracle approaching 227% and Amazon at 125% on a guided basis.
Conclusion
The AI infrastructure boom is creating a depreciation time bomb that’s largely invisible on Big Tech’s income statements. As $695 billion in 2026 capex works its way through 5-6 year depreciation schedules, the impact on operating margins will be substantial and it’s being obscured by accounting practices that spread depreciation across cost of revenue, R&D, and SG&A.
Morgan Stanley estimates that Microsoft, Oracle, Meta, and Alphabet alone could book more than $680 billion in cumulative depreciation charges over the next four years. Michael Burry has called the practice of extending useful life assumptions to flatten this impact “one of the more common frauds of the modern era.”
Whether you agree with that characterization or not, the data is clear: depreciation is the fastest-growing expense line at every major hyperscaler, and it’s hiding in plain sight.
Tools like Daloopa Scout — accessible directly through Claude’s MCP integration — make it possible to cut through the accounting complexity and surface these trends in minutes rather than days. For fundamental analysts and investors, that visibility isn’t optional anymore. It’s essential.
When Amazon announced $200 billion in 2026 capital expenditure guidance on its Q4 2025 earnings call last night, I did what any former buy-sider would do: I opened Daloopa and started pulling comps. Not against other hyperscalers but against the company that actually makes the silicon all that money is chasing- TSMC.
The result was striking. In 2015, the Big 4 hyperscalers (Amazon, Alphabet, Microsoft, and Meta) collectively spent about 3x what TSMC spent on capex. By 2026, that ratio will balloon to nearly 12x. The world’s most important chipmaker is being outspent by its own customers at a pace that is accelerating every quarter and TSMC’s CEO is openly telling you he’s nervous about it.
The Numbers Tell the Story
Using Daloopa, I pulled capex (i.e. PPE) from the cash flow statements of all five companies going back to 2015. For Microsoft, whose fiscal year ends in June, I aggregated quarterly filings into calendar years to create an apples-to-apples comparison.
| Year | Amazon | Alphabet | Microsoft | Meta | Big 4 | TSMC | Ratio |
|---|---|---|---|---|---|---|---|
| 2015 | $4.6B | $9.9B | $6.6B | $2.5B | $23.6B | $8.1B | 2.9x |
| 2016 | $6.7B | $10.2B | $9.1B | $4.5B | $30.5B | $10.2B | 3.0x |
| 2017 | $12.0B | $13.2B | $8.7B | $6.7B | $40.6B | $10.9B | 3.7x |
| 2018 | $13.4B | $25.1B | $14.2B | $13.9B | $66.7B | $10.4B | 6.4x |
| 2019 | $16.9B | $23.5B | $13.5B | $15.1B | $69.1B | $14.9B | 4.6x |
| 2020 | $40.1B | $22.3B | $17.6B | $15.1B | $95.1B | $17.1B | 5.6x |
| 2021 | $61.1B | $24.6B | $23.2B | $18.6B | $127.5B | $30.1B | 4.2x |
| 2022 | $63.6B | $31.5B | $24.8B | $31.4B | $151.3B | $36.3B | 4.2x |
| 2023 | $52.7B | $32.3B | $35.2B | $27.3B | $147.4B | $30.5B | 4.8x |
| 2024 | $83.0B | $52.5B | $55.6B | $37.3B | $228.3B | $29.8B | 7.7x |
| 2025 | $131.8B | $91.4B | $83.1B | $69.7B | $376.0B | $40.9B | 9.2x |
| 2026E | $200.0B | $180.0B | $119.5B | $125.0B | $624.5B | $54.0B | 11.6x |
2026E sources: AMZN $200B (Q4’25 earnings guidance, Feb 6 2026), GOOGL $180B midpoint ($175–185B guidance), MSFT $119.5B (annualized from $29.9B Q4 CY2025), META $125B midpoint ($115–135B guidance incl. finance leases), TSMC $54B midpoint ($52–56B guidance).
Read that last line again. Amazon alone will spend nearly 4x what TSMC plans to spend in 2026. The Big 4 combined will outspend the world’s most critical semiconductor manufacturer by almost 12-to-1.
“I’m Also Very Nervous About It”
What makes this divergence so fascinating is that TSMC’s CEO, C.C. Wei, is essentially telling the market he’s choosing not to match the pace. On TSMC’s Q4 2025 earnings call in January, Wei was remarkably candid when an analyst pressed him on whether AI demand was real:
“You essentially try to ask us whether the AI demand is real or not. I am also very nervous about it. You bet, because we have to invest about $52 billion to $56 billion for the capex. If we did not do it carefully, that would be a big disaster for TSMC for sure.”
— C.C. Wei, TSMC CEO, Q4 2025 Earnings Call
This is the CEO of a company with 95%+ market share in advanced AI chip manufacturing, telling you he’s scared to spend $54 billion. Meanwhile, his customers are collectively deploying $624 billion without blinking.
Wei went on to describe his due diligence process — spending months personally calling cloud service providers to verify their demand signals:
“I talked to those cloud service providers, all of them. I am quite satisfied with the answer. Actually, they showed me evidence that the AI helps their business. I also double checked their financial status: they are very rich… much better than TSMC.”
— C.C. Wei, TSMC CEO, Q4 2025 Earnings Call
Even after satisfying himself that demand is real, Wei emphasized discipline over aggression. On the same call, he stated TSMC would “remain disciplined in our capacity planning approach”, a phrase that would sound prudent in any other context, but takes on a different meaning when your four largest customers are telling you they need more chips than you can physically produce.
The CoWoS Bottleneck: Where the Rubber Meets the Road
The gap between hyperscaler demand and TSMC’s willingness to supply isn’t just theoretical. It’s playing out in real time through TSMC’s advanced packaging technology, CoWoS (Chip-on-Wafer-on-Substrate), which is essential for assembling multi-die AI accelerators that power every data center being built today.
On TSMC’s Q1 2025 call, Wei described CoWoS demand as:
“Almost insane and much, much higher than we can prepare.”
— C.C. Wei, TSMC CEO, Q1 2025 Earnings Call
By Q3, he was still working to “narrow the gap between the demand and supply,” acknowledging that “everything related, like frontend and backend capacity, is very tight.”
Even with TSMC doubling its CoWoS capacity in both 2024 and 2025 and planning to increase monthly output from around 80,000 wafers to as many as 130,000 by the end of 2026, it remains fully booked. NVIDIA alone reportedly secures over 60% of CoWoS capacity, leaving AMD, Google, Amazon, and all other custom ASIC designers competing for the rest. The shortage is so critical that second-tier chip designers have started exploring Intel’s EMIB and Foveros packaging as alternatives, not because they are better, but because CoWoS capacity simply isn’t available.
Why TSMC Won’t Just “Spend More”
There’s a structural reason TSMC can’t or won’t close this gap, and it comes down to the economics of semiconductor manufacturing versus everything else that goes into a data center.
Each dollar TSMC spends goes toward the most capital-intensive, technically complex manufacturing on Earth. Building a single leading-edge fab costs $20 billion or more. The tools alone, particularly ASML’s EUV lithography machines at roughly $380 million each, take years to procure and install. As TSMC’s CFO Wendell Huang noted on the Q4 call, “the capex dollar required to build 1,000 wafers per month of N2 (2nm) capacity is substantially higher than for N3 (3nm).” Each new node is more expensive than the last, with equipment intensity per wafer rising 30–50%.
Meanwhile, hyperscaler capex is increasingly going to everything except chips: land acquisition, building construction, power infrastructure (including nuclear restarts), cooling systems, networking equipment, and electrical grid connections. These are categories where spending can scale much faster than semiconductor fabrication. You can break ground on a new data center campus in months; a new fab takes three to five years from groundbreaking to volume production.
The implication is that TSMC’s share of the total AI infrastructure spend stack is structurally declining. In the 2015–2023 era, TSMC captured roughly 20–25% of the Big 4’s combined capex, suggesting chips were a meaningful share of total infrastructure cost. By 2026, TSMC captures less than 9%. The semiconductor layer is becoming a smaller slice of a much larger pie.
What This Means for Investors
This widening ratio creates several dynamics worth watching.
First, TSMC’s pricing power is immense and increasing. As the sole provider of a critical input with capacity fully booked years in advance, they set the price. Prices for advanced packaging are rising 10–20% annually, and leading-edge wafer prices are following suit. TSMC’s gross margin reached 62.3% in Q4 2025 — a remarkable figure for a capital-intensive manufacturer.
Second, hyperscalers’ capex trajectory may be less about chips and more about other factors. When Amazon guides to $200 billion in capex, much of that is real estate, power, and construction — categories where returns are measured in decades, not quarters. The semiconductor bill, while enormous in absolute terms, is becoming a smaller percentage of the total investment.
Third, the capacity constraint acts as a natural limit on AI expansion. No matter how much money hyperscalers invest in infrastructure, the speed of AI deployment is ultimately limited by how many advanced chips TSMC can produce and package. Wei’s nervousness serves as the market’s safety valve — a disciplined manufacturer unwilling to overbuild helps prevent a repeat of the telecom bubble’s disastrous overcapacity.
As Wei put it:
Can the semiconductor industry perform well for three, four, or five years straight? I’ll tell you the truth, I don’t know. But I look at AI, and it seems like it will be endless… for many years to come.
— C.C. Wei, TSMC CEO, Q4 2025 Earnings Call
That’s the tension at the heart of the AI infrastructure boom: limitless demand meeting deliberately constrained supply, with the ratio between them widening every quarter.
What About Bob’s?: IPO Teardown
Bob’s Discount Furniture (ticker: BOBS) is set to IPO on the NYSE on February 5, 2026. Given the current tariff environment, which is creating meaningful uncertainty for furniture retailers, I wanted to understand their exposure. Using Daloopa’s data sheet combined with Claude’s MCP integration, I pulled together a comprehensive P&L model with a geographic revenue breakdown, sourcing exposure, and a full tariff scenario analysis, all in under 10 minutes, which would normally have taken a day of arduous work.
The Investment Question
Bob’s is a value-focused furniture retailer operating 206 showrooms across 26 states. Their prices are on average 10% below their value-focused peers. The company’s “Everyday Low Prices” strategy depends heavily on imported furniture, thus making tariff policy a critical variable. The S-1 discloses ~73% of cost of goods sold comes from imports. But what does that actually mean for margins under different tariff scenarios?
The P&L- Growth Returns After Weak 2024
Daloopa’s data sheet pulled the historical financials directly from the S-1, organizing quarterly and annual data into a clean model. Here’s what the recent performance looks like:
| Metric | 2023 FY | 2024 FY | 2025 Q3 |
|---|---|---|---|
| Net Revenue ($M) | $2,008 | $2,028 | $1,719 |
| Gross Margin | 46.5% | 46.8% | 45.6% |
| Operating Margin | 5.9% | 5.8% | 6.5% |
| Net Margin | 3.9% | 4.3% | 4.7% |
| Comparable Sales Growth | — | -3.4% | +14.6% |
The story here is solid: comps have turned positive in 2025 (+14.6% in Q3), margins are stable around 46-47% gross, and operating leverage is starting to show. Since 2010, they have posted a 9% CAGR, almost 7% above the home furnishings industry.
Geographic Revenue Mix
Bob’s footprint is concentrated in the Northeast and Midwest, which matters for understanding regional economic sensitivity:
| Region | % of Revenue (2024 FY) |
|---|---|
| New York | 23% |
| Midwest | 23% |
| Mid-Atlantic | 20% |
| New England | 18% |
| West | 16% |
They have 5 distribution centers so expansion likely continues in these areas to reach their target of 500 stores by 2035.
The revenue story is clean and compelling but what about the tariff impact on margin especially given the recent October 2025 on “certain upholstered wooden furniture”?
The Tariff Exposure: This Is Where It Gets Interesting
Using the Daloopa MCP with Claude, I mapped the sourcing exposure based on S-1 disclosures. Having exited China entirely in early 2025, Bob’s now sources from Vietnam, the US, and other Asian countries:
Sourcing Mix (2025)
| Country | % of COGS |
|---|---|
| Vietnam | 63% |
| United States | 27% |
| Other Asian | 10% |
| Total Sourcing | 100% |
Bob’s completed their transition out of China at the start of 2025, significantly reducing their tariff exposure. Now 63% of their sourcing comes from Vietnam, 27% from the US and 10% from other Asian countries such as Malaysia, Thailand and Cambodia. However, the recently imposed tariff on “certain upholstered wooden furniture” adds another layer of complexity. This category-specific tariff started at 25% in October 2025 and is scheduled to increase to 30% in January 2027. It is unclear what these certain furniture items are and whether they will be stacked.
Tariff Scenario Analysis
I modeled three scenarios based on current policy discussions and potential escalation paths:
| Scenario | Vietnam Rate | Others Rate | Rationale |
|---|---|---|---|
| Upside | 10% | 5% | Negotiated reduction |
| Base Case | 25% | 10% | Current trajectory |
| Downside | 46% | 25% | Full escalation |
Quantified Impact
Here’s what each scenario means for Bob’s P&L (based on 2024FY cost structure):
| Scenario | Tariff Cost ($M) | GM Impact (bps) | % of Revenue |
|---|---|---|---|
| Upside | $73M | -360 bps | 3.6% |
| Base Case | $181M | -893 bps | 8.9% |
| Downside | $340M | -1,679 bps | 16.8% |
Key Takeaway: In the base case, tariffs would compress gross margins by approximately 900 basis points, reducing gross profit by roughly a fifth. That’s the difference between a ~47% gross margin and a ~38% gross margin. Significantly better than before their China exit, but still material for a value retailer competing on price.
Investment Implications
- Pricing Power Question: Can Bob’s pass through tariff costs to value-focused consumers? The “Everyday Low Prices” positioning makes this difficult meaning they will likely assume some margin hit as tariffs on Vietnam and other Asian countries evolve.
- Supply Chain Optionality: With 63% Vietnam concentration and complete exit from China, Bob’s has significantly de-risked their supply chain from the most punitive tariff environment.
- Comp Momentum vs. Macro Risk: Strong +14.6% comps show operational execution albeit off a very weak comparable, but tariff headwinds could overwhelm operating improvements.
- IPO Timing: Filing ahead of potential tariff escalation suggests management sees the current window as optimal, which itself is a signal.
Why Daloopa’s Data Sheet and Claude MCP Made This Possible
This analysis would have taken hours to build from scratch, parsing the S-1, structuring the P&L, mapping geographic segments, building sourcing assumptions, and creating the scenario matrix. Daloopa’s data sheet and the Claude MCP integration did the heavy lifting in minutes, pulling structured data directly from the filing and enabling rapid scenario modeling.
To learn more about Daloopa’s MCP, click here.
Disclaimer: This analysis is for informational purposes only and does not constitute investment advice. Country sourcing mix percentages are estimates based on typical furniture retail supply chains and should be validated with company disclosures. Tariff scenarios represent hypothetical policy outcomes.
Motive Technologies filed its S-1 on December 23rd with the formal roadshow kicking off this week. Motive (MTVE on the NYSE) is a software automation platform for the “physical economy,” a dramatically underinvested vertical (only 30% of global venture capital funding since 2015). Their software allows transportation companies to track and improve driver safety and manage fleets more efficiently.
Motive’s 4 core elements are- (1) Driver Safety- GPS devices to increase safety and lower insurance costs. (2) Fleet Tracking/Compliance to meet regulatory requirements. (3) Spend Management and (4) Maintenance to reduce downtime.
I decided to put Daloopa’s new Scout AI Excel agent to the test. Could I build a model for an IPO filing and benchmark it against its publicly traded competitor, Samsara, in the time it used to take just to locate the key data points in the S-1?
The Setup: A Real-Time IPO Analysis Problem
The obvious takeaway is that Motive has solid but hardly spectacular metrics, especially vis-à-vis Samsara. $327.3 million in revenue through Q3’25, up 22% year-over-year. ARR exceeding $500 million up 28% year over year. A 70% gross margin, despite having a hardware segment that basically breaks even. Operating losses are considerable but it is unclear how much was ramped up ahead of the IPO.
Samsara (NYSE: IOT) is the 800-pound gorilla in this space, with $1.75 billion in ARR, 29% growth, and as of Q3 FY2026, 77% gross margins and actual GAAP profitability.
Manually pulling the relevant metrics from Motive’s 200+ page S-1, cross-referencing Samsara’s 10-Q, and building a comparable model would normally take a full day. Scout did it in about 5 minutes.
What Scout Built
I prompted Scout to build a revenue model with key metrics. Then I asked it to run the same framework for Samsara and generate a side-by-side comparison.
Here’s what jumped out:
Revenue & Growth
| Metric | Motive (9M 2025) | Samsara (Q3 FY26) |
|---|---|---|
| Revenue | $327.3M | $416M (quarterly) |
| ARR | ~$500M+ | $1.75B |
| YoY Growth | 22% | 29% |
Samsara is roughly 3.5x Motive’s scale on an ARR basis, growing faster than Motive, albeit decelerating from 35% a year ago to 29% y/y vs. 22% for Motive.
Profitability
Samsara has much higher gross margins at 78% and just reached GAAP profitability in the September quarter, while Motive’s gross and operating margins stand at 70% and low -20%’s, respectively. The focus of the roadshow will clearly be on accelerating revenue growth and its implications for improving cash burn.
| Metric | Motive (9M 2025) | Samsara (9M FY26) |
|---|---|---|
| Gross Margin | 70% | 78% |
| Net Income | $(138.5M) | $(31M) GAAP / Profitable Q3 |
| Operating Margin | Negative | ~19% non-GAAP (Q3) |
Customer Metrics
| Metric | Motive | Samsara |
|---|---|---|
| Total Customers | ~100,000% | Not directly comparable |
| Large Customers ($100K+ ARR) | 494 | 2,990 |
| Large Customer ARR Contribution | 37% | 60% |
| Net Dollar Retention (Large) | 126% | ~115-120% |
| Large Customer Growth YoY | 58% | 30%+ |
Motive is earlier in its enterprise journey, but the trajectory is promising. Large customers grew 58% year-over-year and now represent 37% of ARR, up from 28% a year ago. Samsara’s large customer cohort is more mature, they crossed $1 billion in ARR from $100K+ customers alone this quarter.
The Valuation Question
Samsara trades at roughly 12-13x forward ARR, with a current ~$22 billion market cap. Given the slower growth on a smaller base and cash burn, it will clearly trade at a discount, likely in the mid to high single digits forward ARR, with the discount depending on the messaging from management and Q4’25 results, most importantly 2026 full year guidance.
The Bottom Line
Motive is a legitimate competitor to Samsara with a real product, real customers, and a real growth story. But it’s also a company that needs to prove it can get to profitability without destroying the growth engine. The S-1 gives you all the building blocks to model that path, Scout just removes the friction of actually building the model.
Scout helped me get a quick overview and model structure in minutes, rather than days, allowing me more time to drill into key competitive and margin analyses.
If you’re an existing Daloopa customer, you can request beta access to Scout by clicking here.
The Overreaction
DraftKings (DKNG) and Flutter (FLUT) have been under heavy selling pressure, losing around 25-30% of their market caps in the aftermath of the euphoria around prediction markets’ massive funding and launches. Prediction markets may take away some of the handle from straight sports wagers; however, FLUT and DKNG still have an ironclad hold on the higher-margin parlay handle. Further, prediction markets launching in states where sports betting has not been legalized—avoiding state taxes—will push these states toward capturing lost tax dollars sooner.
We think the market has it backwards.
The Fundamentals Tell a Different Story
Let’s look at what’s actually happening in the business. Both companies continue to grow handle, revenue, and users while prediction markets are supposedly eating into their customer base. Despite downward revisions in guidance due to unfavorable sports outcomes and increased investments, the underlying growth trajectory remains strong, with year-over-year increases in key metrics.
DraftKings Performance
| Quarter | Revenue | YoY Growth | Handle | Structural Sportsbook Hold |
|---|---|---|---|---|
| Q1 2025 | $1.41B | +20% | $13.9B | 10.4% |
| Q2 2025 | $1.5B | +6% | $11.5B | 8.7% (net revenue margin) |
| Q3 2025 | $1.144B | +4% | N/A | N/A |
Full-year 2025 revenue guidance was revised to $5.9B–$6.1B (from an earlier $6.3B–$6.6B), reflecting a 25-30% growth over 2024’s $4.7B, but still indicating resilience amid competitive pressures. User growth continues, with monthly active users contributing to sustained engagement.
FanDuel (Flutter US) Performance
FanDuel, under Flutter, showed robust US performance:
| Quarter | US Revenue Growth (YoY) | AMP Growth (YoY) | Handle | Structural Hold |
|---|---|---|---|---|
| Q1 2025 | +28% (projected full-year to $7.4B) | N/A | $14.6B | 14.5% (Q4 2024 baseline) |
| Q2 2025 | N/A | N/A | N/A | 13.6% |
| Q3 2025 | +9% (group revenue) | +8% (US) | N/A | N/A |
Group revenue for Q1 2025 was $3.7B, with FanDuel driving significant contributions. Full-year group guidance was adjusted to $16.69B due to investments, but US sportsbook AMP grew 5% in Q3, and iGaming AMP surged 30%.
The Parlay Moat Is Real
Here’s what the bears are missing: same-game parlays (SGPs) aren’t just another product—they represent the entire profit engine of modern sportsbooks. While neither company discloses parlay-specific revenue directly, we can see the impact clearly in their margin data. Parlays account for 67-70% of sportsbook revenue and up to 85% of profits for DraftKings and FanDuel, with parlay bets comprising 54-72% of total wagers.
Sportsbook Hold Rates: The Parlay Story
FanDuel’s higher structural margins (e.g., 13.6% in Q2 2025) versus DraftKings’ (e.g., 10.4% in Q1) suggest greater parlay penetration and more efficient promotional spend. If we assume parlays contribute the incremental 5-7 percentage points of margin, that’s roughly 40-50% of sportsbook gross profit coming from parlay-driven hold. The gap between straight bet hold (~5-7%) and structural hold (11-13%) is almost entirely attributable to parlays.
The Liquidity Problem Prediction Markets Can’t Solve
Traditional sportsbooks are principals: they set odds and take the other side of every bet. This means instant execution on any combination of props you want to parlay. Want to bet that Travis Kelce scores a touchdown AND the Chiefs cover the 7-point spread? All the major operators can offer you odds on that instantly and prediction markets may as well with pre-packaged parlays. But what about less common parlays, like: KC -7, Kelce TD, Chris Jones 2+ sacks, and a Mahomes interception?
Prediction markets are exchanges, meaning they match buyers with sellers. Kalshi’s parlay product uses a “request for quotation” model where your custom parlay gets sent out to market makers who decide whether to offer odds.
Think about the combinatorial math: a typical NFL game might have 50+ prop markets. The number of possible 3-leg parlays is 50 × 49 × 48 = 117,600 combinations. For 4-leg parlays? Over 5 million combinations. No market maker is going to provide continuous two-sided liquidity on millions of permutations for every game.
The result? On Kalshi, you wait for a quote that may never come, or you get wider spreads that eliminate the supposed “better odds” advantage. On DraftKings, you click and bet.
This isn’t a technology gap that can be closed with more capital. It’s a structural difference in business models. Being a principal costs more (you bear the risk), but it enables product experiences that exchanges simply cannot match.
The Catalyst: Prediction Markets Will Accelerate State Legalization
This is the contrarian insight: prediction markets aren’t a threat to state legalization, they are the catalyst that finally breaks the logjam in California, Texas, and Georgia.
Consider the political math: State legislators in these markets have been able to kick the can because their constituents had no legal alternative. That’s no longer true. Kalshi, Robinhood, Polymarket, and now DraftKings Predictions and FanDuel Predicts are offering sports event contracts to residents of California, Texas, and Georgia today, with zero tax revenue flowing to state coffers.
The political dynamic has flipped. Before prediction markets, the question was: “Do we want to legalize gambling?” Now the question is: “Do we want to regulate and tax gambling that’s already happening, or cede the market to federal exchanges?”
This is exactly what happened with daily fantasy sports (DFS) in 2015. States rushed to regulate DFS precisely because FanDuel and DraftKings were already operating there. Prediction markets are DFS 2.0, the unregulated activity that forces legislative action. However, this comes amid legal pushback, with at least 9-12 states issuing cease-and-desist orders or lawsuits against platforms like Kalshi, arguing they violate state gambling laws.
DraftKings and FanDuel Are Playing Both Sides
Here’s what the market is missing about DraftKings’ and FanDuel’s prediction market launches: they’re not abandoning their sportsbooks, they’re creating an on-ramp for future customers.
DraftKings Predictions launched on December 19, 2025, in 38 states, including California, Texas, and Georgia. FanDuel Predicts launched on December 22, 2025, initially in five states with a phased national rollout through early 2026. Users in non-legalized states will download the apps, create accounts, deposit funds, and start betting on event contracts. When those states eventually legalize traditional sports betting, DraftKings and FanDuel already have the customer relationships, apps installed, and payment credentials on file.
This is the daily fantasy playbook all over again. DFS was never a great standalone business—it was a customer acquisition channel for sports betting. Prediction markets are the new DFS: a way to build brand awareness and customer databases in non-legal states.
Valuation: Paying for a Problem That May Not Exist
DraftKings trades at roughly 18x 2026 while Flutter trades at 12x multiple on 2026 adjusted EBITDA, respectively. DKNG is a pure play on the US sports betting and iGaming market while FLUT has a broad geographic composition. Both stocks are down 25-30% from their recent highs on prediction market fears.
But the prediction market “threat” is being priced as if:
- Kalshi can replicate the full sportsbook product experience (they can’t—see parlays).
- States will never legalize (the opposite is more likely now).
- DraftKings and FanDuel will lose customers (they’re still growing).
- The regulatory environment will remain favorable to prediction markets (multiple states are suing, and legal battles are ongoing).
The asymmetry here is favorable. If prediction markets are banned or restricted, these stocks re-rate higher immediately. If prediction markets succeed in forcing California/Texas legalization, DraftKings and FanDuel get access to 70 million new customers. If the status quo persists, both companies continue growing their existing markets while building customer bases via their own prediction market products.
The Bottom Line
Wall Street is pricing DraftKings and FanDuel as if prediction markets are an existential threat. The reality is more nuanced: prediction markets are structurally unable to compete on the highest-margin products (parlays), they’re creating political pressure that will accelerate state legalization, and the incumbents are already adapting by launching their own prediction offerings.
We tested leading AI agent frameworks on 500 financial questions. The best reached 90% accuracy. For production finance work, that’s not enough.
The promise and the gap
AI agents that can autonomously search, reason, and retrieve data represent a significant step forward for financial research. Instead of manually navigating databases and documents, an analyst could ask a question and receive a verified answer in seconds. We focused on single-number lookup—questions like “How many trucks did Volvo deliver in Asia in Q4?” We wanted to measure how close we are to that reality. To find out, we built an evaluation framework and tested three frontier agent systems: OpenAI’s Agents SDK with GPT-5.2, Anthropic’s Agent SDK with Claude Opus 4.5, and Google’s ADK with Gemini 3 Pro. Each agent faced 500 questions in two configurations: web search with reasoning enabled, and adding Daloopa’s structured database via MCP. This study extends our earlier work from consumer chatbots to production agent frameworks, with root-cause analysis for every failure.Main finding: The 90% ceiling
The best configurations reached approximately 90% accuracy. All three frontier models converged to similar performance when given the same tools.| Configuration | Accuracy |
|---|---|
| Claude Opus 4.5 (+Daloopa MCP) | 90.8% |
| Gemini 3 Pro (+Daloopa MCP) | 90.6% |
| GPT-5.2 (+Daloopa MCP) | 89.2% |
- For every 10 data points retrieved, expect 1 to be wrong
- A 50-company financial screen would contain roughly 5 incorrect values
- These aren’t random typos—they’re systematic misinterpretations that a human reviewer might not catch
US versus non-US: A clue in the data
Before diving into the failures, we noticed a pattern. All three models performed better on US companies than non-US companies:| Model | US Accuracy | Non-US Accuracy | Gap |
|---|---|---|---|
| Claude Opus 4.5 (+Daloopa MCP) | 93.4% | 87.8% | +5.6pp |
| Gemini 3 Pro (+Daloopa MCP) | 92.6% | 88.2% | +4.4pp |
| GPT-5.2 (+Daloopa MCP) | 90.4% | 87.8% | +2.6pp |
Why the 10% fails
We analyzed every failure from the best-performing configuration (Claude Opus 4.5 +Daloopa MCP, 46 incorrect answers). The errors fall into distinct categories.Period confusion: 63% of errors
Nearly two-thirds of all failures stem from a single pattern: fiscal versus calendar period confusion. When asked for data from “fiscal year ended March 2023,” agents queried2023FY. But Daloopa’s database uses the starting-year convention, so the correct query was 2022FY. The agent retrieved data from the wrong year entirely.
This isn’t a reasoning failure. The agent understood the question correctly. It failed because the tool’s period naming convention wasn’t documented clearly enough for the model to infer the correct mapping.
The pattern is systematic and fixable. Better tool documentation and client-side prompt engineering could address nearly two-thirds of all current errors.
Wrong series selection: 20% of errors
Financial databases contain many similarly-named series. Agents sometimes selected the wrong one:- “Gross production” instead of “net production”
- A sub-component instead of the total
- A broader category instead of a specific line item
The best AI with databases was worst at web search
The most surprising finding wasn’t which model performed best. It was this paradox:Not all web search is the same
Different AI platforms have different web browsing capabilities. Claude’s web search tool returns search result snippets—it can see previews but can’t read full pages or PDFs. Google and OpenAI’s tools can browse documents directly and extract data from tables. It’s the difference between having library catalog access and actually reading the books.Finding answers isn’t the same as trusting them
Even when Claude found relevant information, it often couldn’t commit. In 55% of its WebSearch failures, Claude located a plausible answer but declined to provide it—continuing to search until timing out or explicitly giving up. In one case, Claude searched for Alcoa’s maintenance cost forecast. It found “$10 million” in a search result—the correct answer. But instead of stopping, it ran five more searches looking for “confirmation,” and eventually concluded: “I could not locate the specific dollar amount.” The answer was there. The tool couldn’t verify it confidently.The insight
AI reliability depends on both the model and the tools it’s given. With MCP, all three frontier models converged to 89-91% accuracy. Without it, tool quality dominated—and the “best” model fell to last place. Structured data levels the playing field. But as the 90% ceiling shows, it’s necessary, not sufficient.The path to 99%
Based on our error analysis, improving from 90% to 99% requires work across multiple dimensions:| Improvement | Estimated Impact |
|---|---|
| Better tool documentation (period conventions) | Addresses 63% of errors |
| Improved series disambiguation | Addresses 20% of errors |
| Client-side prompt engineering | Compound benefits |
| Data quality improvements | Contributes to ~7% of errors |
Study design and limitations
We generated 500 single-number financial retrieval questions across six categories: income statement, balance sheet, cash flow, guidance, operational KPIs, and segment data. Each question has a verified ground-truth answer from company filings. Each agent was tested in two configurations, both with reasoning enabled:- WebOnly: Web search only, no structured data access
- +Daloopa MCP: Web search plus Daloopa’s financial database
Conclusion
Frontier AI agents with structured data access reach approximately 90% accuracy on financial data retrieval. That’s notable progress, but not yet reliable enough for unsupervised production use. The errors aren’t random. Nearly two-thirds stem from a single fixable pattern: fiscal period naming conventions. Another fifth come from ambiguous series selection. These are infrastructure problems, not fundamental limitations of the models. The path to 99% is visible:- Better tool documentation that makes conventions explicit
- Improved data disambiguation at the source
- Client-side prompt engineering for edge cases
- Continuous evaluation to measure each improvement
For software stocks- gross margins, ARR growth rate, GRR, NRR and free cash flow margins are the key metrics but with CRE its quite different. Real estate companies are valued on FFO and AFFO, why you ask? Lets dig in.
As public companies, commercial real estate companies need to follow GAAP guidelines that require them to depreciate their assets over their useful lives (i.e. 39-40 years for commercial buildings) which has a large drag on reported eps. So unlike GPUs who lose value yearly when the new version is faster and more efficient, prime real estate generally appreciates in value.
In order to get an accurate view of normal operating cash flow, the two key metrics used are FFO and AFFO. Funds from Operations (FFO) adds back the depreciation and amortization related to the real estate and then removes gains/losses from property sales because they are non-recurring. Adjusted Funds from Operations (AFFO) goes even further to compute the true operating cash flow from the business by deducting the normal capital expenditures needed to maintain the properties. AFFO also smooths out rent adjustments.
The Challenge
I wanted to quickly analyze the major players across CRE subsectors. Traditionally, this means pulling 10Qs, normalizing metrics across different reporting conventions, and manually building a comp table.
Building the Model with Scout
Scout let me pull together a comprehensive valuation comparison in minutes. I started with five REITs spanning different CRE segments. Prologis (Industrial/Logistics), Realty Income (Triple-Net Retail) and Simon Property Group (Regional malls) are more traditional commercial real estate companies while Equinix and Digital Realty operate data centers.
For each company, I pulled TTM metrics across the key REIT-specific KPIs: FFO per share, AFFO per share, EPS, EBITDA, and dividends. Scout normalized everything to trailing twelve months automatically, summing the last four quarters (2024Q4 through 2025Q3) without requiring manual adjustments.
What the Data Shows
The resulting model surfaces some interesting disparities:
Valuation spreads are wide. Price/AFFO ranges from mid-teens for Realty Income and SPG while the data center companies are in low twenty times. Prologis trades at the highest relative AFFO multiple, will need to drill in there to see why the hefty premium. Traditional retail-oriented REITs are trading at meaningful discounts to data center plays no surprise given AI infrastructure demand, but the spread is notable.
Dividend yields vary significantly. Realty Income leads at 5.3%, while Equinix yields just 2.6%. For income-focused investors, the yield pickup in traditional REITs versus data centers is substantial.
Balance sheet quality differs. Debt/EBITDA of 4.7x at Prologis versus 5.6x at Realty Income shows the range of leverage across the group. Lower leverage provides more flexibility if rates stay elevated longer than expected.
Why This Matters
The macro setup is compelling: CRE valuations at multi-decade relative lows, potential Fed easing on the horizon, and divergent fundamentals across subsectors. Having standardized, quarterly data across companies lets you move quickly from thesis to actionable analysis.
What would have taken hours of manual work involving downloading filings, finding the right line items, normalizing for fiscal year differences, Scout compressed into minutes. The output is a clean, auditable model I can update each quarter as new data flows in. Scout also allows adding other companies and metrics to the model by directing it in the prompt via the add-in.
For anyone evaluating CRE exposure, whether you are overweight data centers and curious about rotating into traditional REITs, or underweight the sector entirely and looking for entry points, having this kind of rapid industry modeling capability changes the workflow entirely.
As AI requires large quantities of data—and success in financial services is dependent upon having trusted data—the limitations of public, web-sourced inputs can be glaring. Daloopa closes this gap by delivering the most accurate and complete data infrastructure. The platform covers 5,000+ public companies globally, delivers up to 10 times more data points per company than other providers, and each datapoint is hyperlinked to its original source for transparency and auditability.
Daloopa provides the core AI data infrastructure that powers financial agents and is trusted by the leading global AI companies. Its MCP also powers analytical AI workflows ranging from hedge funds identifying quarter-over-quarter inflections and simulating scenarios, to equity researchers creating reports with full source traceability.
“We are excited to continue executing on our strategy of being the data infrastructure for AI and agentic workflows in financial services. Daloopa is the core foundation of today’s AI-enabled research stack. Building a connector with ChatGPT is a tremendous opportunity to expand access to high-quality fundamental data and transform the financial landscape,” said Thomas Li, CEO of Daloopa.
Already integrated with Anthropic’s Claude for Financial Services, Daloopa’s MCP is LLM-agnostic and supports Claude, OpenAI APIs, and other AI platforms using MCP Standard Protocol.
About Daloopa
Daloopa is the financial data layer powering the finance ecosystem with the most accurate and comprehensive data. Its proprietary platform sources, structures, and distributes this historical financial dataset covering 5,000+ public companies globally. Analysts at the world’s top investment firms trust Daloopa’s workflow solutions to save valuable time and accelerate their decision making. Daloopa also provides the critical AI data infrastructure that underpins the best financial agents and is trusted by the world’s most preeminent AI companies.
For more information or to request a demo, click here.
Media Contacts:
Christina Scott
Vice President of Marketing
[email protected]
Today, Goldman Sachs downgraded DLTR from Buy to Sell due to concerns that the low-income consumer is showing material weakness. HundredX’s data shows that the purchase intentions of low-income households, defined as those with incomes between $25,000 and $40,000, declined noticeably in October, matching the annual low seen in April. Using Daloopa’s MCP, Dollar General (DG) was identified as the most exposed among the dollar stores, much more exposed than DLTR. FIVE screened as the least exposed and possibly a good pair trade against DG.
https://claude.ai/share/f0d1bf0b-bb95-4bc0-a2cb-251a48ed8be3
While all 3 dollar stores screen as having high exposure to this cohort, DG screens as the highest with a 95% rating. Over 82% of their Q2 sales came from the consumables category. DLTR has significantly less exposure to consumables at 50% and has less rural penetration compared to DG. FIVE screens as the least exposed of the three as they benefit from teen/youth spend being less cyclical and a middle-income customer buffer.
To learn more about Daloopa MCP, visit: https://test-wp.daloopa.com/products/mcp
]]>As a former equity portfolio manager, I’ve learned that the best opportunities often lie behind the most obvious headwinds. When everyone is trying to quantify further downside and give up on an idea, my contrarian instincts kick in. Hershey (HSY) presents exactly this kind of setup today—a quality business temporarily hit hard by unprecedented commodity inflation, creating what could be a textbook value recovery story.
Getting Up to Speed Fast: The Power of Daloopa + Claude
As an investor with mainly a technology background, I needed to quickly understand the key factors driving HSY, especially how cocoa price fluctuations had affected their business over the last five years. Instead of spending hours analyzing earnings transcripts and building a model from scratch, I used Daloopa’s new MCP (Model Context Protocol) integration with Claude to evaluate the company’s financial performance.
Within minutes, I had pulled comprehensive data showing Hershey’s remarkable resilience through the cocoa crisis, with hyperlinks taking me exactly to the filing where these numbers were pulled from:
Hershey’s Performance Through the Cocoa Storm (2020-2024)
| Year | Net Sales | Gross Margin | Net Income | EPS | Cocoa Derivative Impact |
| 2020 | $8.15B | 45.4% | $1.28B | $6.31 | +$28.9M cost |
| 2021 | $8.97B | 45.1% | $1.48B | $7.34 | +$78.8M cost |
| 2022 | $10.42B | 43.2% | $1.65B | $8.22 | +$40.8M cost |
| 2023 | $11.16B | 44.8% | $1.86B | $9.31 | -$97.7M gain |
| 2024 | $11.20B | 47.3% | $2.22B | $11.22 | -$563.0M gain |
What jumped out immediately was the company’s incredible operational performance despite the commodity headwinds as they effectively hedged the cocoa risk. Revenue grew 37% over five years, EPS increased 78%, and remarkably, gross margins actually expanded to their highest levels in 2024. However, 2025 has not been so kind to Hershey…
The Perfect Storm Creates the Perfect Setup
2025: Peak Pessimism
This year has been brutal for Hershey’s financials, but that’s exactly what creates the opportunity. Using Daloopa’s real-time data, I could see the dramatic impact:
- Q1 2025: Commodity and derivatives costs hit $618.9—the highest on record, resulting from higher commodity costs and derivatives mark-to-market.
- Q2 2025: Additional $592M in cocoa-specific cost increases
- Full Year 2025: Adjusted EPS expected to drop 36-38% to $5.54-$5.86 range
Management lowered full-year 2025 EPS guidance after Q2 to reflect an increased tariff hit of $170-180 million, up from the $20 million initially estimated. Management stated that they would have increased the EPS guidance, save for the additional tariff headwind. CEO Michele Buck said talks with the Trump administration indicated some willingness to exempt natural resources that cannot be produced in the US.
The Catalyst Everyone’s Missing
While investors focus on 2025’s depressed earnings, they’re overlooking the most significant development: Hershey announced a “low double-digit” price increase across its confectionery portfolio on their July 23rd earnings call. Note, it takes around 90 days for the price increases to take effect so Q4 will be the first quarter to really see the impact. This isn’t just any pricing action—it represents roughly 16 points of pricing contribution on 80% of their business.
The Math That Gets Me Excited
- Pricing Benefit: ~$1.4B incremental revenue from 16-point increase
- Margin Expansion: Management targets 500+ basis points of gross margin recovery
- EPS Normalization: From ~$5.70 in 2025 to $10+ potential in 2026, Street is well below that in the $7.00 range but a steady increase throughout the year would be ideal for the stock.
- Cocoa Stabilization: Prices have already retreated from $12,000/MT peaks to ~$7,500/MT
- Buyback Resumes: HSY has not repurchased shares since the March 2024 quarter as they preserved cash due to the cocoa cost crisis. As EBITDA/FCF increases in 2026, the debt to EBITDA of 1.9x and heading lower should support a resumption of the share buyback.
The Daloopa Advantage: Speed to Insight
What would have taken me hours of manual data collection and analysis, Daloopa’s MCP integration with Claude accomplished in minutes. I could quickly:
- Pull five years of financial fundamentals with proper citations
- Analyze derivative impacts and commodity hedging effectiveness
- Compare pricing actions across different periods
- Model various scenario outcomes with real-time data
This isn’t just about convenience—it’s about competitive advantage. In today’s markets, the first analyst to identify inflection points wins.
Risk Management: What Could Go Wrong
Key Risks:
- Volume elasticity exceeds management’s 1:1 assumption
- Cocoa prices spike again (though supply forecasts look better)
- Consumer spending weakens, pressuring premium pricing
- Tariffs escalate beyond the current $170-180M annual impact
Risk Mitigation:
- Management has shown exceptional hedging sophistication ($563M gain in 2024)
- A strong balance sheet provides flexibility
- Diversification into salty snacks reduces cocoa dependency
- Brand moats historically support premium pricing
The Setup: Asymmetric Risk/Reward
At current levels, HSY offers what I look for in quality value plays:
Downside Protection:
- Trading below historical multiples despite stronger fundamentals
- Dividend aristocrat with 55-year track record
- Defensive characteristics of food staples
Upside Catalysts:
- Margin recovery as pricing flows through
- Potential cocoa cost normalization
- Tariff relief possibilities
- Market re-rating as earnings inflect higher
- Share buyback resumption as credit metrics improve
The Bottom Line
Hershey represents exactly the type of opportunity I built my career finding: a quality business temporarily impaired by cyclical headwinds, trading at depressed multiples with clear catalysts for recovery. The market is pricing in permanent damage to a company that just delivered solid performance during the worst commodity crisis in decades.
Sometimes the best investments are hiding in plain sight, wrapped in temporary bad news. HSY might just be one of those times.
Disclosure: This analysis is for educational purposes only and does not constitute investment advice. Always consult with a qualified financial advisor before making investment decisions. Data sourced from Daloopa.